Pearl River County Flood Plain Management Project, Pearl River County, MS
Contract No. DACW38-02-D-0002
Task Order No. 0006
PROJECT REPORT
1. DESCRIPTION OF WORK (from the USACE Scope of Work).
USACE, Vicksburg District requires the acquisition of LIDAR elevation data in Pearl River County Mississippi. The project shall include capture of the LIDAR data as well as the creation of digital terrain models derived from the data set. The Contractor shall collect digital elevation data from a precision airborne (LIDAR) survey within the entire project area. The proposed project area is approximately 530,000 acres. The purpose of the survey is to obtain measurements of the bare ground surface, as well as top surface feature elevation data, for providing geometry input to USACE hydraulic models. The survey will be accomplished utilizing aerial topographic Light Detection and Ranging (LIDAR) technology to develop digital terrain models and obtain elevations of the surface features and the bare ground.
2. GROUND CONTROL SURVEY.
Ground Control was done in two phases. The first phase was done by Pyburn and Odom under contract to USACE, Vicksburg District. Forty-three (43) permanent concrete markers with Vicksburg brass caps on 18in copper-weld rods were set throughout Pearl River County, MS. Pyburn and Odom performed a GPS survey to establish horizontal and vertical control on these monuments using NGS control monuments published on the NAD 83 & NAVD 88 Datum.
The second phase was done by Dungan Engineering under contract to Pearl River County. Dungan Engineering provided the necessary material and staff for the targeting and the maintenance of 40 aerial targets for and during the aerial campaign of the referenced project. In an amendment to the contract, Dungan Engineering provided the necessary equipment, material and staff for the establishment and maintenance of 23 additional aerial targets for and during the aerial campaign of the referenced project to control an additional flight for 200-scale orthophotography.
3. LiDAR FLIGHT.
a. Flight Planning. TrackAir v.1.99H was used to determine the optimal flight plan for the project. The following were incorporated into the planning:
Repetition Rate / 10 KHzScan Frequency / 13 Hz
Altitude (AGL) / 4,100’
Scan Angle (Full FOV) / 40 Degrees
Side Overlap / 30%
Mean X Spacing / 7.739’
Mean Y Spacing / 7.762’
Foot Print / 1.23’
b. Acquisition.
1. Three GPS base stations supplied and operated by Sea Systems Corporation were used to support precise positioning and orientation of the ALTM’s sensor head during the entire duration of flight. The GPS base stations were Trimble 5700 receiver units utilizing Zephyr Geodetic antennas. Each GPS base station was located within the boundary of the project area.
2. The initial aerial survey was completed over the course of 8 days. Data collection started around 23h30 UTC on Saturday, 01 February 2003. Flightlines completed during this flight were lines one through 12. On 01 February the flight commenced at 02h50 UTC and completed lines thirteen through twenty-nine. The flight on 02 February began around 23h10 UTC and collected lines thirty through thirty-eight. A second flight was then flown beginning around 02h30 UTC on 03 February and completing lines thirty-nine through forty-five. On 04 February the flight commenced around 22h40 UTC and covered lines forty-five through fifty-four. The second flight followed a refueling stop around 02h30 UTC and completed lines fifty-five through sixty-six. The flight on 05 February covering lines 67-69 and 97 through 100 began around 22h10 UTC and ended around 00h30 due to weather. The final day of initial data collection occurred on 08 February. Two flights were flown this day. The initial flight began around 00h46 UTC and covered lines seventy through eighty-eight and line 103. The second flight began around 22h19 UTC and completed lines 67-69, 89-96 and 101 and 102. This completed the initial LIDAR data collection for the project and the ground crews continued in their remaining work in and around the project area. The aircraft and personnel involved during the LIDAR portion of the survey were demobilized on the night of Sunday, 09 Feb 2003.
3. Following a preliminary examination of the collected data it was determined that one flight was required to refly some of the collected lines. A Cessna Skymaster 337, N111AT, was mobilized from Huntsville International Airport, Huntsville, AL to Picayune Municipal Airport, Picayune, MS on 30 Mar 2003. This aircraft was outfitted with an Optech ALTM 1210 LIDAR system. Data collection commenced at approximately 22h35 UTC and constituted reflying lines 5, 9, 86-88, 92 and 103 for various technical reasons. This completed the LIDAR data collection for the project and the ground crews continued in their remaining work in and around the project area. The aircraft and personnel involved during the LIDAR portion of the survey were demobilized on Monday, 31 Mar 2003.
4. PHOTO FLIGHT.
1. Flight Planning. a. Flight Planning. TrackAir v.1.99H was used to determine the optimal flight plans for the project. The following were incorporated into the planning:
100-Scale Ortho Flight Plan ParametersAltitude (AGL) / 4,800’
Forward Overlap / 80%
Side Overlap / 30%
200-Scale Ortho Flight Plan Parameters
Altitude (AGL) / 9,030’
Forward Overlap / 60%
Side Overlap / 30%
400-Scale Ortho Flight Plan Parameters
Altitude / 12,000’
Forward Overlap / 80%
Side Overlap / 30%
b. Acquisition.
1. Three GPS base stations supplied and operated by Sea Systems Corporation were used to support precise positioning and orientation of the photo centers during the entire duration of flight. Each GPS base station was located within the boundary of the project area.
2. The aerial survey was completed over the course of 3 days. Data collection started around 11h19 local on Tuesday, 11 February 2003. Flightlines completed on this day ranged from one to nine at 4800 feet and one through five at 9030 feet. Collection recommenced around 9h47 local on 12 February. Lines completed during this flight were six through 12 at 9030. On 13 February collection began around 09h26 local and lasting through 15h00 local. Lines collected during this flight included ten to eighteen at 9030 and ten through twenty-three at 12000 feet. This completed the photo collection for the project and the ground crews continued in their remaining work in and around the project area. The aircraft and personnel involved during the photo portion of the survey were demobilized during the afternoon of Thursday, 13 February 2003.
3. Upon inspection of the film it was determined that reflights would be necessary. On 23 February 2003 a Cessna 335, N918AA, was mobilized from Huntsville International Airport, Huntsville, AL to Picayune Municipal Airport, Picayune, MS outfitted with a RC30 Camera and AGFA Pan 80 film. Collection took place between 09h34 and 12h31 local. Lines six, eight and nine at 9030 and lines sixteen, seventeen, twenty and twenty-three at 12000 were collected.
5. AGPS/IMU PROCESSING.
Upon completion of the flight portions of the project the GPS data was post processed for quality and backed up. For redundancy and accuracy purposes, the airborne GPS data were processed from the base stations using GrafNav from Waypoint Consulting, Inc. These trajectories were used in the processing of the inertial data. The inertial data were processed using PosProc from Applanix, Inc. This software produces an SBET (“smooth best estimate of trajectory”) using the GPS trajectory from GrafNav and the roll, pitch and heading information recorded by the POS (Position and Orientation System). Results were favorable for all flights and errors are estimated to be less than 5cm.
6. PHOTO PROCESSING.
The aerial film was processed and edited at the Atlantic photo lab. The film was titled according to the Vicksburg District specification and one set of black and white contact prints produced. A total of 1,074 prints at all three scales were shipped to the Vicksburg District on March 31, 2003.
7. SCANNING.
Scanning was done at 14μm for all three scales of photography. This resulted in the following approximate resolutions:
PHOTO SCALE / SCANRESOLUTION / TARGET ORTHO RESOLUTION
1”=800’ / 0.441’ / 0.5’
1”=1500’ / 0.827’ / 1.0’
1”=2000’ / 1.102’ / 2.0’
8. LiDAR DATA PROCESSING.
Data collection of the survey areas resulted in a total of 103 flight lines covering the project area including 3 control lines. The tapes, flight logs, raw air and ground GPS files were then taken to the office for data processing using Realm, a LiDAR processing software package from Optech. Processing began by downloading these files into a Realm database.
Realm uses the SBET to generate a set of XYZ data points for each laser return. Data can be segregated based on the first- and last-pulse information. First and last pulse files were created during the processing of this dataset. This project’s data were processed in strip form, meaning each flight line was processed independently. Processing the lines individually provides the data analyst with the ability to QC the overlap between lines.
Raw LiDAR data are processed within the LiDAR manufacturer’s software to produce XYZI files. These files are output in UTM coordinates with a corresponding Ellipsoid Height value. Output XYZI files from Realm were converted from UTM co-ordinates with GRS80 ellipsoid elevations into State Plane Coordinate System (NAD83) with NGVD29 orthometric heights using the U.S. Army Corps of Engineers’ Corpscon, version 5.11.08. Corpscon utilizes the Geoid96 model for the ellipsoid to orthometric height conversions. The resultant XYZI files were subsequently imported into a project, on a per pulse basis, using TerraScan (Terrasolid Ltd.) where each line was checked against adjacent lines.
This check revealed an issue with the calibration numbers being used for the system. Further investigation led to the understanding that calibration parameters would have to be determined on a line-by-line basis. Though uncommon, this situation is not unheard of with airborne laser terrain mapper systems. The system parameters were determined through a number of iterations involving looking at the calibration line flown for the flight, comparisons between the survey line and the control lines as well as interline tie comparisons.
Once the calibration parameters for each line were determined and the data recalculated, the data was checked against the control and validation points across the project area. The results of these checks showed a bias in the dataset for all lines, save for 97 and 99, of -1.2 U.S. Survey Feet. It was determined that an adjustment to correct for this bias would be best for the dataset. A subsequent check of the DEM found it fitting the validation and control points well.
The data from each line was then combined and a classification routine performed to determine the rough surface model. This initial surface model was then reduced using MD Atlantic’s proprietary methods to create the final bare-earth dataset. A Triangular Irregular Network (TIN) was generated using the final surface data. Contours were then created from the TIN for use in performing a quality control of the surface. The LiDAR data were taken into a stereo environment and melded with photogrammetric data. Breaklines were subsequently compiled along hydro features to support the contour generation.
9. AERIAL TRIANGULATION.
Point selection, coding and measuring was performed on all three scales of photography using Z/I ISAT. Block adjustments were made using PhotoT, ver 4 (within ISAT).
The objective of the AT was to control the photos utilizing the provided 63 ground control points. GPS airborne was also used in the solution. This was accomplished by performing a QA/QC of the provided AGPS data utilizing automatically generated photogrammetric tie (parallax) points. The ground control targets were then measured,
and combined with the AGPS and tie points. The results were then analyzed by orienting the stereo models, and measuring elevations in the project area for proper leveling and positioning.
The results achieved from the solutions of all three scales will support the production of orthophotography to ASPRS Class II standards.
As the requirement for Orthophotography was eliminated from the Vicksburg District task order and subsequently contracted to Atlantic by Pearl River County, the report and associated files were delivered to the Pearl River County Addressing Office.
10. ORTHOPHOTOGRAPHY.
Production of the 100 & 400-scales of orthophotography was done by MD Atlantic. The 200-scale orthophotography was done by Michael Baker Corporation.
Ground pixel resolution for the three scales are as follows:
PHOTO SCALE / ORTHO SCALE / RESOLUTION1”=800’ / 1”=100’ / 0.5’
1”=1500’ / 1”=200’ / 1.0’
1”=2000’ / 1”=400’ / 2.0’
As this was eliminated from the USACE task order and subsequently contracted to MD Atlantic by Pearl River County, delivery of the orthos was made to the County. A copy of all three scales of orthos is included on the final LiDAR data delivery to USACE.
11. DEM PRODUCTION.
The final data from the LiDAR classification and breakline collection were imported into ArcInfo to generate TIN models. The points and breaklines were used as input in the ArcInfo TIN creation. The TINs were clipped to correspond with the 400-scale ortho tile index.
Once the TINs were created, the ArcInfo command TINLATTICE was used to create GRIDs with a 5’ posting.
12. MISCELLANEOUS SHAPEFILE PRODUCTION.
Shapefiles of the contours were created using SDS-FIE standards. Two files were delivered; 2’ contours within the two incorporated areas of the county and 5’ contours of the entire county. These files are attributed in accordance with SDS-FIE standards.
A shapefile was generated for the breaklines and flight lines. The breakline 3D shapefile attributes are generic with no other attributes added. The flight lines are attributed with flight line number.
Point shapefiles were created for the survey data. Three files were delivered; control, calibration and validation. The control shape consists of all aerial targets and published control utilized in the network adjustment. It is attributed with Easting, Northing, Z and point name. The calibration shape contains the RTK data from the local calibration site established. It is attributed with Easting, Northing, Z and description. The validation shape contains RTK data of the vegetation categories required to be measured for LiDAR verification. It is attributed with Easting, Northing, Z and description.